I have been forced to think about deposition issues lately, because a) depositing a structure, and b) refereeing a paper which discusses PDB problems..
The GUIs in use in both CCP4 and Phenix offer tools to help the depositor but it is perhaps too easy to now deposit without due care and attention.. Errors I have made - 1) first download of files is covered by the GUI and that triggers a validation report.. 2) try to correct the structure according to that report.. 3) decide to reprocess the data to a lower? higher? resolution - use different software? etc, then rerefine against that data, but fail to re-upload structure factors. or 4) correct the model , maybe add an extra water, et etc then upload that file without repeating the refinement Then of course there is the hands-on approach where the depositor makes their own stupid errors - mismatch of information which one hopes the PDB checks will flag.. But re RSRZ outliers - I don't know how these are generated. Looking at my latest attempt - a reasonable structure at 1.7A - quite a few apparently acceptable LEU VAL ALA - even a beautiful TRP - are flagged with poor RSRZ , whereas the floppy high Bfactor ARG and LYS whose positions are at best an intelligent guess are not.. Is the score based on matching the expectation using the refined B values to the actual density? Eleanor On Thu, 30 Jan 2025 at 17:47, Clemens Vonrhein < 0000daef624adb06-dmarc-requ...@jiscmail.ac.uk> wrote: > Dear all, > > That's the kind of interesting question that triggers me to have a > closer look at PDB entries ... just to convince myself that a value of > 30% RSRZ outliers really is as disturbing as I initially thought. > > One can go through all available validation reports of X-Ray PDB > structures and extract the > > _pdbx_vrpt_summary_diffraction.percent_RSRZ_outliers > > value. This gives the attached plot percent_RSRZ_outliers.png (with > the same plot also on logscale and as a close-up to zoom into the >20% > range). > > * Looking at it his way, only 0.32% of current PDB entries have a > RSRZ outlier of 30% or higher. > > Unless one has e.g. a 3-residue small peptide structure with one > poor residue I would be very reluctant to stick with the model/data > in its current shape ;-) > > * Most of those very high values seem very suspect and are most > likely due to problems with the reflection data (or the use of it > in the validation software), i.e. deposited CIF file is > corrupted compared to the original file a user tried to deposit: > > * out of the 193 structures with a value of 50% or higher, there > are 8 structures with a reported resolution below 4A > > * but there are also 70 structures of 1.5A or better > > * 8 of the apparently "worst" structures (100% RSRZ outliers) have > a reported resolution of 1.32-0.96A ... unlikely that those > density-fit parts of the validation reports are correctly > reflecting the model [1]. > > Not sure that helps much, but it satisfied my curiosity at least :-) > > Cheers > > Clemens > > [1] Looking at those high-resolution, poor density fit structures is > quite entertaining: some have all density shifted (maybe deposited > the wrong reflection data for the correct model, which can happen > for large group depositions). Others have some really dodgy > look-n-feel on the PDBpeep server that analyses the deposited > data. > > > On Wed, Jan 29, 2025 at 11:43:15PM +0800, 白雪慧 wrote: > > Happy New Year, everyone. > > > > Recently, I have encountered some problems when submitting PDB, and I > have seen some discussions about these problems before. > > > > Meanwhile, I still want to ask how to reduce the RSRZ outliers in the > wwPDB validation report, is there any good way to reduce the RSRZ outliers > > > > in the WWPDB Validation report? I enclose my alidation report. > > > > ######################################################################## > > > > To unsubscribe from the CCP4BB list, click the following link: > > https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=CCP4BB&A=1 > > > > This message was issued to members of www.jiscmail.ac.uk/CCP4BB, a > mailing list hosted by www.jiscmail.ac.uk, terms & conditions are > available at https://www.jiscmail.ac.uk/policyandsecurity/ > > > > -- > > *-------------------------------------------------------------- > * Clemens Vonrhein, Ph.D. vonrhein AT GlobalPhasing DOT com > * Global Phasing Ltd., Sheraton House, Castle Park > * Cambridge CB3 0AX, UK www.globalphasing.com > *-------------------------------------------------------------- > > ######################################################################## > > To unsubscribe from the CCP4BB list, click the following link: > https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=CCP4BB&A=1 > > This message was issued to members of www.jiscmail.ac.uk/CCP4BB, a > mailing list hosted by www.jiscmail.ac.uk, terms & conditions are > available at https://www.jiscmail.ac.uk/policyandsecurity/ > ######################################################################## To unsubscribe from the CCP4BB list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=CCP4BB&A=1 This message was issued to members of www.jiscmail.ac.uk/CCP4BB, a mailing list hosted by www.jiscmail.ac.uk, terms & conditions are available at https://www.jiscmail.ac.uk/policyandsecurity/
